Building Energy prediction has emerged as an active research area due to its potential in improving energy efficiency in building energy management systems. Essentially, building energy prediction belongs to the time series forecasting or regression problem, and data-driven methods have drawn more attention recently due to their powerful ability to model complex relationships without expert knowledge. Among those methods, artificial neural networks (ANNs) have proven to be one of the most suitable and potential approaches with the rapid development of deep learning. This survey focuses on the studies using ANNs for building energy prediction and provides a bibliometric analysis by selecting 324 related publications in the recent five years. This survey is the first review article to summarize the details and applications of twelve ANN architectures in building energy prediction. Moreover, we discuss three open issues and main challenges in building energy prediction using ANNs regarding choosing ANN architecture, improving prediction performance, and dealing with the lack of building energy data. This survey aims at giving researchers a comprehensive introduction to ANNs for building energy prediction and investigating the future research directions when they attempt to implement ANNs to predict building energy demand or consumption.@en

Space cooling in buildings is responsible for massive energy consumption and carbon emissions. Accurate cooling load prediction can facilitate the implementation of energy-efficiency cooling control strategies in practice. In this paper, an improved attention-based deep learning approach is proposed for robust ultra-short-term cooling load prediction. First, a novel time representation learning is introduced to extract the periodicity and non-periodicity of cooling loads efficiently. Then, long short-term memory with an attention mechanism extracts properly the time steps by identifying the relevant hidden states and learns high-level temporal dependency. The approach additionally incorporates extreme gradient boosting through the error reciprocal method, enhancing the elimination of prediction errors and improving robustness. The study takes Guangzhou as an example and generates cooling loads using diverse occupancy schedules of five building types based on the Chinese National Standard and Typical Meteorological Year data. The approach is evaluated on datasets comprising the cooling loads, meteorological data, and contextual information. Through results analysis, the approach outperforms other models in terms of prediction accuracy and robustness across all building types. Additionally, model interpretation is provided regarding feature importance and attention matrixes, which enhances the understanding and transparency of the final prediction from the proposed approach.@en

The energy matching of PV driven air conditioners is influenced by building load demand and PV generation. Merely increasing energy performance of building or PV capacity separately may improve the energy balance on a large time resolution, the real-time energy mismatching problem is still serious. In this study, a coordinated optimization method of PV capacity, building design, and load flexibility is proposed for improving the real-time energy matching of PVAC system. Then, a methodology integrating data mining method (XG Boost) and parametric simulation was developed to identify the determinant parameters of PV system and building design, exploring feature importance and correlations. The results of XG Boost indicate that the PV capacity, shape factor, and SHGC are the most critical factors. Finally, based on the optimized building design, the PCM layer was applied to improve the real time energy matching. To achieve a goal of 90 % ZEP, the PCM capacity can be decreased by 50.4 % and 62.8 % in Guangzhou and Shanghai in the optimized building. Moreover, the PV capacity can be reduced by 23 % in Guangzhou. The findings of this study provide practical guidance for designing PVAC system coupling with building design and energy storage devices.@en

Real-time nonintrusive occupancy estimation can maximize the use of existing sensors to infer occupant information in buildings with the advantages of fewer privacy concerns and fewer extra device costs. Recently, many deep learning architectures have proven effective in estimating occupancy directly from raw sensor data. However, some handcrafted features manually extracted from statistical and temporal domains might convey additional information for occupancy estimation. In this study, a novel knowledge fusion network for nonintrusive occupancy estimation is proposed to integrate knowledge from two streams, i.e. automatic knowledge stream from a deep learning architecture and handcrafted knowledge stream from manual feature engineering. Moreover, four different fusion modules are investigated to optimize the design of the fusion network. To verify the effectiveness of the proposed network, experiments are conducted in a dataset from the ASHRAE Global Occupant Behavior Database, which is collected from an office space with records of indoor environment parameters, occupant-building interactions, and contextual information. The results demonstrate the superiority of the proposed fusion network, which outperforms five representative algorithms. Furthermore, the ablation study underscores the benefits of knowledge fusion and occupant-building interaction information, showing that the proposed fusion network can enhance the occupancy estimation accuracy by 3.47 % to 9.24 %.@en

Integrating renewable energy is a promising solution for buildings to achieve the net-zero-energy goal. Expanding real-time matching between renewable energy generation and building energy demand can help realize more enormous zero-energy potential in practice. However, there are few studies to investigate the real-time energy matching in renewable energy building design. Therefore, in this study, a hybrid ensemble learning framework is proposed for analyzing and predicting zero-energy potential in the real-time matching of photovoltaic direct-driven air conditioner (PVAC) systems. First, the datasets of zero-energy probability (ZEP) are generated under the three main climate regions in China, which are with consideration of the load flexibility of air conditioners and based on six important design variables. Second, a novel ensemble learning method named Extreme Gradient Boosting (XGBoost) is selected to predict ZEP and the Bayesian Optimization (BO) is adopted to identify the optimal hyperparameters and further improve the prediction performance. The statistical analysis shows that ZEP distributions are very different from one region to another one and the PVAC systems in Beijing are the easiest to achieve the zero-energy goal. Among all the variables, PV capacity is the most significant and positively related to ZEP. The prediction results show BO-XGBoost achieves more than 99% accuracy and outperforms other benchmark models in the ZEP prediction of three cities. In a word, this paper reveals BO-XGBoost is the most effective model for ZEP prediction and provides the framework for designers to utilize zero-energy potential analysis and prediction for the first time.@en

Reducing the heating and cooling load through energy-efficient building design can help decarbonize the building sector. Heating and cooling load prediction using machine learning (ML) techniques become increasingly important in the rapid assessment of building design variables at the early design stage. However, when applying the ML techniques, it still requires expert knowledge and manually frequent intervention to improve the prediction performance. Hence, this study proposed an automated machine learning (AutoML)-based framework to automatically generate the optimal ML pipelines for heating and cooling load prediction. An experimental dataset of residential buildings was used to evaluate the proposed framework. The proposed framework achieved the best performance with R2 of 0.9965 and RMSE of 0.602 kWh/m2 for heating load prediction, and R2 of 0.9899 and RMSE of 0.973 kWh/m2 for cooling load prediction. The prediction results showed that the proposed framework outperformed the other improved ML models from the representative studies in the last five years. Further, an explainable analysis of the ML models was explored to reveal the relationships between design variables and heating and cooling load. The proposed framework aims at promoting the AutoML-based framework to designers for building energy performance prediction without excessive ML knowledge and manually frequent intervention.@en

A general and simple semianalytical method based on the Galerkin procedure is introduced in this letter for wave propagation problems related to radially inhomogeneous cylindrical dielectric waveguides with arbitrary permittivity profiles. The complicated wave propagation problem in an inhomogeneous waveguide is transformed into a simple problem of solving an equivalent matrix in terms of the Galerkin procedure. Eigenvalue solutions are given for three kinds of representative radially inhomogeneous cylinders to verify the generality of the proposed semianalytical method. The accuracy and the improvement of the convergency are illustrated by comparing the calculated results with those obtained by the analytical method or the matrix numerical method.@en

The real-time energy matching between building load and PV generation is low in actual applications of photovoltaic direct-driven air conditioners (PVACs). The indoor thermal comfort temperature range (TCTR) can enhance the load flexibility of PVACs to improve the real-time zero energy probability. Therefore, a real-time zero-energy potential evaluation method with the adoption of TCTR for PVACs was proposed. The indoor temperatures are mainly determined by real-time PV generation, which minimizes the power taken from the grid and batteries. The indoor temperature, conditioned by PVACs under varying operating conditions, is predicted using machine learning models at a one-minute time resolution. The real-time energy matching performances of PVACs are evaluated by key indicators, including zero-energy probability, real-time zero-energy probability, complete overcooling rate, and complete overheating rate, for different climatic regions. With a fixed indoor setting temperature in summer, the real-time zero-energy probabilities in Shanghai, Guangzhou, and Beijing are 1.89%, 2.45%, and 2.11%, respectively. While adopting the TCTR with the similar PV capacity, the corresponding values in Shanghai, Guangzhou, and Beijing reach 53.13%, 49.48%, and 87.11%, respectively. In addition, the real-time zero-energy points frequently appear when large cooling demands are needed. Finally, an optimization of PV capacity is conducted. The optimal PV capacity was found to be at the point where seasonal PV generation is 1.3 times the dominant seasonal energy consumption. PVACs with TCTR can utilize the PV generation and load flexibility to the greatest extent and the evaluation method for PVACs is beneficial for designing a more practical and economical system.@en

This study investigates the diagnostic capabilities of a Diagnostic Bayesian Network (DBN) for air handling unit (AHU) components, particularly focusing on the heat recovery wheel (HRW) and heating coil valve (HCV). Unlike data-driven methods relying heavily on high-quality labeled data, this knowledge-based DBN is more suitable for real-world applications, where labeled faulty and normal data are hard to obtain. Notably, existing studies predominantly concentrate on developing DBN for AHU with recirculated air, neglecting thorough investigations into AHU with HRW, a prevalent system in North Europe and increasingly recommended post-COVID-19 for mitigating viral propagation. This paper presents a DBN setup with expert knowledge for an AHU with HRW, which is evaluated using experimental data from an office building in the Netherlands. The results show that the proposed DBN can successfully diagnose typical faults in HRW and HCV. @en